Multi-stage quantitative risk assessment of a critical system in mining industry

1. Introduction

Engineering Asset Management (EAM) is a systematic approach towards managing and maintaining physical assets throughout their lifecycle. EAM plays a crucial role in ensuring optimal performance, reliability, and longevity of physical assets. It integrates engineering, financial, and operational strategies to enhance asset reliability and performance while minimizing risks and costs. Proper asset management ensures that machinery, equipment, and infrastructure operate at peak efficiency, minimizing failures and maximizing productivity. A study by Amadi-Echendu, et al. [1] highlights that asset management plays a vital role in mitigating operational risks and ensuring compliance with safety and environmental laws. According to ISO 55000 standards [2], sustainable asset management supports efficient resource utilization, waste reduction, and compliance with environmental policies. Asset Management (AM) has traditionally been regarded as a routine discipline. However, it should now be seen as a strategic philosophy that must be integrated across all levels of an industry. By fostering awareness throughout the industry, AM can enhance the understanding of the importance of optimizing engineering asset performance, ensuring alignment between industry objectives and asset management goals [3].

The industry consists of different types of equipment that work together synchronously to form a network of equipment, which can deliver the desired output with maximum efficiency. To maintain maximum equipment efficiency, it is important to understand the criticality of every network element. Criticality is a means of assessing the effect of each failure mode within the equipment portfolio and assessing their associated risks. Criticality analysis can be an essential step in reliability engineering, providing a systematic way to identify high-risk assets and prioritize maintenance efforts. The key factors which might help to assess the criticality are failure likelihood, failure consequences, interdependency of the systems and sub-systems, cost implications and associated risks.

Risk is a very subjective topic and a very important at the same time. The understanding of the risk of a failure mode depends on the perspective of the observer which can also influence the criticality of that failure mode. Risk can be driven by one or multiple defined factors and must be assessed against these to determine the level of focus necessary to manage it. Risk management includes the application of logical and systematic methods for establishing the process of identifying, analysing, evaluating, treating, monitoring, reviewing, reporting, and recording risks. Risk assessment is that part of risk management which will help to identify and analyse the criticality of a failure mode and its consequences [4]. Risk assessment involves identifying, evaluating, and communicating the presence, characteristics, extent, frequency, influencing factors, and uncertainties associated with potential losses [5]. It has been considered a powerful approach to address public concerns and to develop sound policy and design strategy [6]. Several benchmark studies have shown that risk assessment results might differ depending on the industry professionals performing the analysis and the clients requesting them [7]. This subjectivity in the assessment process raises the question of uncertainties. Also, the future development and maintenance of the infrastructure of society will even more likely demand an intensified and quantified focus on risk [7]. Thus, due to tremendous demand of risk-based decision-making in engineering applications and a significant lack of recognition of risk analysis as a necessary discipline, there has been development of a range of practices for risk analysis [7]. These practices have been published and well defined by author Valis and Koucky [4].

In this paper we have proposed a novel means of applying common Failure Mode and Effects Analysis (FMEA)/ Failure Mode, Effects, and Criticality Analysis (FMECA) – centred tools to identify ‘critical’ process systems. This multi-layered risk assessment practice will be applied to a critical system to evaluate the influence of individual equipment failure risks within it. The paper uses a case study data for implementing the novel approach of risk assessment and compare various risk assessment methods, and parameters evaluated for the implementation of multi-stage risk assessment framework. The paper critically reviews all the results derived through different implemented risk assessment methods in this paper and showcases the advantages and limitations of the proposed novel approach.

2. Review of FMEA and FMECA methods

The advancement of mining technology has led to the creation of complex technical systems that require a systematic analytical and methodological approach to be properly understood. These systems have emerged due to the increasing demand for and interest in resource extraction. Effective risk analysis and management play a crucial role in ensuring quality and reliability within the mining industry. However, one of the main challenges in technical systems is conducting thorough risk assessments. Historically, risk management in mining has not been given sufficient attention, but there is now an urgent need for change [8]. A key approach to risk analysis involves proactive error identification through methods such as FMEA and FMECA. These methodologies break down systems at a functional level, considering failure modes as the loss of specific component functions.

The concept of FMEA originated in the 1940s within the U.S. military and was later adapted by the aerospace, automotive, and manufacturing industries [9]. FMEA is a methodology designed to identify the ways in which components, systems, or processes can fail. FMEA identifies all the potential failure modes associated with core function/s of a system or process and their effects respectively. It also evaluates the risk of each failure mode that may disrupt any such function. A failure mode is the observed reason of failure or the reason of incorrect performance [4]. To avoid failures above a nominated risk threshold, appropriate corrective actions are drafted as outputs for implementation. As per McDermott, et al. [10], FMEA has evolved into a widely used technique for improving quality and minimizing failures in both product and process design. Generally, FMEA’s are performed during the design or process development of a greenfield project or modification stages.

The applications of FMEA can be found across many industries like machinery [11-14], electronics [15], chemical [16], medicine [17, 18], textile [19, 20], aerospace, nuclear, mining and other manufacturing industries [8, 21-23]. Focusing more closely on the mining industry as an application, these practices demand strong, decision-making criteria to help define site-relevant factors that influence the detectability, severity and occurrence of each failure mode; many of which are complex and multi-dimensional in nature. These criteria help to define and understand the availability, reliability, maintainability and overall criticality of the systems. In the mining industry, this decision-making process is challenging and complex, as failure of a single item of equipment can lead to a sudden or delayed system-wide stoppage. Furthermore, buffers may also exist between adjacent systems (for example, stockpiles) which make the actual effect of a failure difficult to determine across the broader process at large. In such cases, decisions are likely focused around the system’s availability, with maintenance strategy improvement and subsequent reliability gains acting as controls to mitigate associated risks. Currently, in the mining industry decision-making processes for asset investment are largely qualitative, with assets being monitored and managed via qualitative FMEA tools or other risk-based methods [3]. Author Duda and Juzek [24] illustrates the application of FMEA method for hazard identification and process risk assessment in a coal mine which is a crucial step in risk assessment and safety management, helping to prevent accidents and ensure operational reliability. Paper [25] highlights the use of FMEA to overcome the uncertainty in the decision-making process of an underground coal mine. Most recent application of FMEA method is highlighted in a case study on risk analysis of a machine breakdown in a cement factory in Indonesia [26].

When used as a top-down analysis tool, FMEA may only identify the most significant failure modes within a system. However, when applied as a bottom-up approach, FMEA can complement Fault Tree Analysis method by uncovering additional causes and failure modes that contribute to top-level system issues. Despite its widespread adoption, traditional FMEA has limitations, including subjectivity in RPN scoring (different assessors may assign different scores), lack of consideration for failure interdependencies and inability to predict unknown failure modes. It cannot effectively detect complex failure modes involving multiple failures within a subsystem or predict the failure intervals of specific failure modes at higher system levels. Additionally, the method of calculating the Risk Priority Number (RPN) by multiplying severity, occurrence, and detection rankings can lead to inconsistencies, sometimes assigning a higher RPN to a less critical failure mode than a more severe one. This issue arises because these rankings use an ordinal scale, where the numerical values indicate relative order but not precise magnitude. For example, a ranking of “2” is not necessarily twice as severe as a ranking of “1”, nor is an “8” necessarily twice as bad as a “4”. However, multiplication treats them as if they are proportional, leading to misleading prioritization [27].

To address these limitations, modern approaches like FMECA offers an extension to FMEA theory, allowing each failure mode to be ranked according to its importance or criticality. This critical analysis is usually qualitative or semi-quantitative but may be quantified using actual failure rates [27]. The FMECA is the result of two steps: – Failure Mode and Effect Analysis (FMEA) and Criticality Analysis (CA). The CA can be aligned with two distinct alternatives: qualitatively or quantitatively. The implementation of the quantitative analysis method is well explained by author Lipol and Haq [27]. FMEA and FMECA are methodologies used to identify potential failures in a product or process. While both follow the same fundamental approach, they differ in key aspects. FMEA provides qualitative insights, whereas FMECA incorporates some quantitative data that can be measured. FMEA is commonly used in industries as a “what-if” analysis tool and is an integral part of NASA’s flight assurance program for spacecraft. On the other hand, FMECA assigns a criticality level to failure modes and is used by the U.S. Army to evaluate mission-critical equipment and systems [27]. FMECA is essentially an extended version of FMEA. To conduct FMECA, analysts first perform an FMEA and then carry out a CA. FMEA identifies failure modes and their effects, while CA ranks these failures based on their severity and occurrence rate, prioritizing the most critical ones [27].

The most recent application of FMECA can be found in the field of medicine [28, 29], manufacturing [30-34], machinery [35-38], aerospace [39, 40], and automobile [41]. Cheng, et al. [42] applied FMECA to assess the reliability of a metro door system. They compiled failure statistics for various metro door subsystems and determined their criticality levels. The analysis identified the EDCU function as the most critical subsystem. Ćatić, et al. [43] conducted a criticality analysis of the steering tie-rod joint. They first created a layout of the joint using a tree diagram, outlining its various components. Subsequently, they performed a Fault Tree Analysis (FTA) for the steering tie-rod joint. Focusing on the recent applications of FMECA in mining industry, paper [44] successfully assessed the criticality of dumper subsystems using the FMECA methodology. Criticality indices for each failure mode were determined based on failure rate, frequency, and operating time. The impact of different operating time modes on criticality rankings was also analysed. Among the eight dumper subsystems, the engine component was identified as the most critical, ranking first. Additionally, the relationship between failure occurrences and criticality was examined and validated using Spearman’s correlation test, confirming that higher failure occurrence values correspond to greater criticality. Paper [45] highlights the application of FMECA to assess and design the reliability of the coal system in the Oslomej surface mine. This approach helped identify potential failure modes, enhance reliability evaluation, and conduct a qualitative criticality analysis. As a result, key insights were gained, guiding attention toward the highest-risk areas. Paper [46] introduces a method for supporting occupational risk management in quarry blasting operations using a modified FMECA algorithm. The proposed approach systematically identifies risks and highlights key occupational hazards that should be prioritized for preventive measures. These preventive actions can be incorporated during the design phase by modifying technology or work organization, depending on the specific quarry's available options. Table 1 summarises the reviewed literature and their corresponding research gaps, and highlights how the proposed research offers improvements over existing approaches.

While the reviewed literature provides valuable contributions, the comparison in Table 1 underscores a clear and recurring gap across existing studies. These gaps collectively point to the need for a more comprehensive and data-driven approach to maintenance decision-making, one that aligns more closely with the operational realities of the mining industry. This broader need forms the foundation and motivation for the present study. As mining operations grow increasingly complex and cost-sensitive, the ability to make informed, data-driven maintenance decisions has become more critical than ever. Traditional tools such as FMEA and FMECA, while widely adopted, often rely on static and qualitative assessments that do not fully capture the dynamic and economic nature of equipment failures. There is a growing need for approaches that can translate real-world failure behaviour into actionable insights for minimizing downtime, reducing costs, and improving asset reliability. This study responds to that need by proposing a novel multi-stage quantitative risk assessment framework centred around a unique application of conventional FMEA/FMECA theory. This practice will be used to compare the influence of equipment failure modes on the overall criticality of the system in which they function. The multi-stage quantitative risk assessment framework assesses failure rate (likelihood), downtime and cost (Consequence) as quantitative elements in the FMECA framework. The work is focused on the mining industry and the data for the study was sourced from a gold mining company in Australia.

3. Data

The work presented in the paper is based on the case study of a gold mining company in Australia. The industry failure data was collected from two different sources, one providing maintenance work order information and another providing downtime information. These were labelled ‘Selective work orders.xlsx’ and ‘Downtime.xlsx’ respectively. The raw data was recorded manually over different periods for different systems, downloaded in comma-separated value (CSV) format, and was initially analysed in MS Excel. Through this process, a new dataset was created which was largely tailored to this undertaking. The initial exploratory analysis using information originally from the ‘Downtime.xlsx’ client spreadsheet. The major focus was on extracting variables like failure modes, downtime associated with those failure modes and finally the downtime cost associated with each failure mode. As the industry manually records the data, it was difficult to find consistency in the data across the time frames as highlighted in Table 2.

The identified critical system based on downtime from Table 2 is Modular Crusher as highlighted in Fig. 1, with its year-wise downtime bifurcation. The downtime history of individual systems presented in Fig. 1 are in minutes.

Fig. 1Analysis of system downtime

Table 3 shows the different failure modes of all the equipment associated with the Modular Crusher system.

Some commonly observed failure modes were blockage of jaws, damaged belt, bogged conveyor, bearing failure, liners/bolts failure, oil and lubrication issues, chute issues, electrical issues, etc.

However, ‘Total Circuit’ does not resemble any equipment but rather it was a qualitative choice made by the operator to record a downtime event. for instance, if a conveyor bearing fails and interrupts the entire circuit, some operators recorded this as a ‘Total Circuit’ failure. from an asset criticality perspective, it had to be assigned to ‘conveyor’ otherwise it had no use. Some of the undefined failure modes were listed under ‘others’, and were omitted as they could not be processed.

4. Comparison of risk assessment frameworks

Using a dataset derived from the equipment listed in Table 3, outcomes from two, industry-adopted qualitative risk assessment processes will be compared against a novel, multi-stage quantitative approach. Firstly, risk assessment outcomes forwarded by traditional FMEA practices will be shown. These will be followed by outcomes from a common industry-based method. Lastly, a novel, multi-stage quantitative FMECA practice will be introduced for comparison. Not only will this highlight the influence of ‘subjectivity’ in the repeatability of qualitative risk assessment outcomes, but it will also help us understand the concept of risk by looking through different lenses.

5. Comparison of the three methods

The three different approaches explained in this paper yields three different results. to understand these better, this section, compares each of these approaches by focusing on a specific critical equipment, the ‘Screen’. Table 10 displays the standard FMEA process results. as mentioned earlier in Section 4.1, the RPN is the product of the occurrence (O), severity (S) and detection (D) of a failure (RPN = O * S * D). in this example, we will focus on the occurrence (O) and severity (S) to measure the likelihood and consequence of the failure mode respectively. Further to this, detectability is an additional measure associated with the ‘monitoring effectiveness’ of a failure mode. It requires an intricate knowledge of the maintenance strategies that safeguard the failure mode to be applied consistently. This factor has been omitted from this study due to this ‘intricate knowledge’ being largely unavailable. if the data is more granular and if more strategy information is available then ‘detectability’ should be included, unfortunately that is the constraint of this study.

According to the 1st person, risk-prone failure modes include ‘Electrical fault’ and ‘Mechanical fault’ which have been assigned an RPN of 90. Conversely, the 2nd and 3rd identify the most risk-prone failure modes to be ‘conveyor bogged’ and ‘belt damaged’ respectively. Along with the inconsistency in identifying the most risk-prone failure mode, there is also widespread variation when the RPN value is ranked highest to lowest in each case. This indicates that there is significant uncertainty in the maintenance decision-making process.

Table 11 highlights the application of the company’s generic process to the same piece of critical equipment, the ‘Screen’. According to generic approach the most risk-prone is ‘bearings failure’ with risk score as E25.

Before explaining the proposed approach of this paper, which is known as multi-stage quantitative FMECA, the different parameters of this framework are analysed. This is important for understanding, as there are specific factors that are assessed individually using a more defined classification system. This helps to significantly improve decision-making uncertainty from the previous approach.

Turning our attention to the novel approach, the failure modes will be assessed quantitatively using two parameters. the first being ‘failure rate’, which is a ‘likelihood’ measure, and the second being ‘downtime’, which is often a dominant measure for ‘consequence’ in the context of equipment failure modes. Initially, the failure rate of all failure modes in the example will be calculated, followed by its corresponding downtime in hours. Table 12 and Table 13 display the corresponding results.

Thus, in the first stage, ‘Mechanical fault’ failure mode is of high priority with 33 failure occurrences over an annualised period. This corresponds to an ‘Almost Certain’ rating when the company’s generic (industry-aligned) risk likelihood matrix is applied.

If the company’s generic (industry-aligned) consequence matrix is applied, a ‘high’ consequence rating is appropriate to this failure mode given it has amassed a total of 312 hours of downtime per event. as highlighted in Fig. 2, the top three critical failure modes based on their likelihood are ‘Mechanical fault’, ‘Maintenance activities’ and ‘bogged conveyor’.

If a decision-maker had to rely on just the likelihood score, then the most frequently occurring failure mode will be attended first. in this case the ‘Mechanical fault’ followed by the other in sequential order. But if a different person tries to analyse the same failure modes through the lens of ‘downtime’ alone, the priority changes. as seen in the Fig. 3, the topmost critical failure modes coincidently reoccur, but this time they are reordered, “mechanical fault”, “bearing failure” and then “maintenance activities”. Thus, both the parameters yield different priorities thus creating uncertainties in the decision-making process.

Fig. 2Analysis of failure rate (likelihood)

Fig. 3Analysis of failure modes of an identified critical equipment

After this individual analysis of this parameters, Table 14 showcases their role in the novel quantitative approach proposed in this paper.

This approach reveals why it is important to consider both ‘failure rate’ and ‘downtime’ to highlight the likelihood and consequences of the risk associated with an event caused by each particular failure mode. It is important to understand that we are analysing historical data, and we are analysing risk per event as we do not want another event to occur. Risk is an event-based measure, and a quality decision-making process must therefore assess the risk associated with each unique event-type.

The higher the frequency of failure in a particular period of time then it is more likely that event is to occur again. The risk matrix (Likelihood x Consequence) presented in this framework indicates the risk per event. The downtime parameter is important, but it doesn’t belong in likelihood space. Only the failure rate belongs in the likelihood space because its matrix describes classifies the time period between failure events due to a specific failure mode. Similarly, the downtime is important when we are discussing consequence. The consequence is described by the cost parameter because the cost is inherently related to the production downtime per event, contributed by the same failure mode.

This framework generates a risk score which is basically dependent on the total loss per event (Consequence). The total loss is basically the downtime in hours due to each occurred event which is then multiplied by the production loss cost per hour. Thus, by using Table 5, 6 and 7 we can evaluate the risk per failure mode, per event. The risk score of the ‘Bearing failure’ event is ‘E20’ which is a high priority considering the consequence of that failure mode. By comparing it with the individual parameter analysis; with respect to failure rate displayed in Table 12, a ‘Bearing failure’ was not even in the top three in the priority list, but with respect to downtime it would have been addressed, but not as a high priority. This shows how much the cost parameter makes difference in the perspective of decision-makers, as it results in the development of more robust maintenance strategies and utilises funds in a more efficient and effective way.

The findings of this research contribute meaningfully to both theoretical development and practical decision-making in the context of maintenance management within the mining industry. The proposed multi-stage risk assessment approach based on quantitative FMECA framework introduces a novel way to assess failure criticality by incorporating failure rate, downtime, and cost into a dynamic and repeatable analysis process. This approach moves beyond the limitations of traditional RPN-based methods, which often rely on static, subjective evaluations, and instead supports a continuous reassessment of risk as operational conditions evolve. From a theoretical perspective, this research enhances the existing body of work on reliability-centered maintenance by integrating data-driven, multi-criteria decision-making into the FMECA process. The inclusion of time-sensitive and cost-related parameters makes the framework especially relevant to industries like mining, where equipment downtime directly impacts productivity and revenue. On the managerial side, the framework provides maintenance planners and operations managers with a transparent and systematic tool to prioritise interventions based on measurable outcomes. By linking technical degradation with economic impact, it facilitates more accurate forecasting, targeted resource allocation, and improved maintenance planning. This alignment between technical risk and financial consequence also enables better justification of maintenance strategies to senior management, reinforcing accountability and supporting long-term asset performance optimisation.

6. Conclusions

The aim of every industry is to reduce the risk associated with their assets operation. the risk generally increases with the ageing of the assets and every industry has a different approach towards assessing this risk. Assessing risk is a crucial phase of the maintenance decision-making process and it is very important to complete this process as consistently as possible. There are different qualitative and quantitative risk assessment techniques adopted throughout industry to help reduce uncertainty. Each of these techniques are often universally applied to equipment and systems; regardless of criticality. Every technique has some level of uncertainty or limitation. for example, different likelihood and consequence factors may be relevant to different organisations, or, different organisations may record events differently or inadequately, which may influence the risk assessment approach that can be applied. Without question, it can be stated that a lack of quality data increases the challenges of the risk assessment process.

Risk assessment is a very time consuming, and a costly process and every industry adopts, to some extent, a fixed annual budget for asset maintenance and improvement activities. in such scenarios, industries go for the most affordable and quick options to arrive at a decision, which is not always a recommended approach. This is a widespread problem that is a result of pace of development, production targets and ever-changing market demands. in this style of workplace, it is very important to equip the decision-making process or the decision-maker with multiple decision parameters which can be selected based on data that is readily available throughout industry at a relatively high quality (i.e. availability, downtime and budget figures).

The aim of this paper was to design a framework based on multiple quantitative parameters. in this paper, a novel quantitative risk assessment approach has been introduced. This approach is based on quantitative FMECA. The entire framework is designed with different decision-making parameters, represented as a standalone individual decision-making practice. This gives flexibility to the decision-making committee to plan the maintenance activities in accordance with time and budget availability. The novel risk assessment approach considers assessing, analysing and prioritising the failure modes using a multi-stage, quantitative approach. Whilst it is important to note that all three approaches yield results that will reduce overall risk, the cost and timeframe needed to do so will vary dramatically. Ultimately, the management team can select an option that better suits their needs and maturity. However, this paper has determined that a quantitative influence is required to streamline decision-making processes and reduce outcome subjectivity.

By breaking down risk assessment into distinct stages, this methodology addresses key limitations of traditional FMEA/FMECA. The inclusion of failure rate in the first stage helps to quantify the likelihood of failures occurring. The second stage, which assesses downtime, allows for a more comprehensive understanding of how failures impact system availability and overall productivity. Finally, integrating cost analysis in the third stage provides financial justification for prioritizing maintenance and reliability improvements. This multi-dimensional approach ensures a balanced decision-making process that aligns with both technical performance and economic feasibility. Furthermore, applying multi-stage quantitative FMECA in industries such as mining, manufacturing, and aerospace can significantly reduce unplanned downtime, optimize maintenance planning, and enhance system reliability. As industries continue to adopt data analytics and predictive maintenance, this approach can be further refined by integrating machine learning algorithms and real-time monitoring to improve risk prediction accuracy.

In conclusion, the multi-stage quantitative FMECA methodology presented in this paper represents a critical advancement in reliability engineering and risk management. By systematically analysing failure rate, downtime, and cost, organizations can enhance safety, improve asset performance, and reduce operational losses, leading to more efficient and sustainable industrial operations. It also shows how important it is to filter the failure modes through an economic lens along with other quantitative risk assessment parameters which yield results in a more effective and efficient way.